Journal of Materials Processing Technology 210 (2010) 14361454

Contents lists available at ScienceDirect

Journal of Materials Processing Technology
j our nal homepage: www. el sevi er . com/ l ocat e/ j mat pr ot ec
History of plasticity and metal forming analysis

Kozo Osakada

Osaka University, School of Engineering Science, Machikaneyama 3, Toyonaka, Osaka, Japan

a r t i c l e i n f o
Article history:
Received 6 November 2009
Received in revised form 1 April 2010
Accepted 1 April 2010
Keywords:
Metal Forming
Plasticity
History
a b s t r a c t
The research history of mechanics, physics and metallurgy of plastic deformation, and the development
of metal forming analysis are reviewed. The experimental observations of plastic deformation and metal
forming started in France by Coulomb and Tresca. In the early 20th century, fundamental investigation
into plasticity ourished in Germany under the leadership of Prandtl, but many researchers moved out
to the USA and UK when Hitler came in power. In the second half of the 20th century, some analyzing
methods of metal forming processes were developed and installed onto computers as software, and they
are effectively used all over the world.
2010 Elsevier B.V. All rights reserved.
1. Introduction
The phenomenon of plasticity has been studied from the view
points of mechanics, physics andmetallurgy, andmanymathemati-
cians contributedto rene the mechanics of plasticity. The research
results are applied to geophysics and strength of materials, and
of course, are used as indispensable tools for analyzing the metal
forming processes.
Although the theories and the experimental results are
explained in many text books, the building up process is not known
well. Inthis article, thehistoryof plasticityinrelationtotheanalysis
of metal forming is reviewed by putting emphasis on the personal
prole of researchers.
2. Strength of materials and plasticity before the 20th
century
2.1. Early days of strength of materials (Timoshenko, 1953)
Leonardo da Vinci (14521519) left many texts and sketches
related to science and technology although he did not write books.
One of the examples he studied is the strength of iron wire, on

This paper was originally presented as a keynote at the 9th International Con-
ference on Technology of Plasticity, hosted by the Korean Society for Technology
of Plasticity, under the Chairmanship of Professor D.Y. Yang. The paper is a unique
record of the history of a key area of interest for the Journal of Materials Processing
Technology, so we invited Professor Osakada to expand the paper and submit it to
the Journal. We are extremely grateful for the permission of Professor Yang to allow
us to present the paper here. J.M. Allwood and A.E. Tekkaya, April 2010.

Tel.: +81 78 841 2594; fax: +81 78 841 2594.

E-mail address: k.osakada@hm.h555.net.
which hangs a basket being lled with sand. The strength of the
wire can be determined by measuring the weight of sand when
the wire is broken. Unfortunately, the idea and the advancement
made by da Vinci were buried in his note, and were not noticed by
scientists and engineers of the time.
It is generally accepted that Galileo Galilei (15641642) is the
originator of modern mechanics. In his famous book Two NewSci-
ences, he treatedvarious problems relatedtomechanics, including
an example of the strength of a stone beam. He organised his
methods applicable to stress analysis into a logical sequence. His
lecture delivered in the University of Padua attracted many schol-
ars gathered from all over Europe, and disseminated the methods
of modern science.
Robert Hooke (16351704) published the book Of Spring in
1678 showing that the degree of elongation of the spring is in pro-
portion to the applied load for various cases. It is generally believed
that Hooke came up with the idea of elastic deformation when he
carried out experiments on the compressibility of air at Oxford
University as an assistant of Robert Boyle (16271691), who put
forward Boyless law.
2.2. Torsion test of iron wire by Coulomb (Bell, 1984)
In the paper submitted to the French Academy of Sciences in
1784, C.A. de Coulomb showed the results of torsion tests of iron
wire carriedout withthe simple device giveninFig. 1. He estimated
the elastic shearing modulus of the material from the frequency of
torsional vibration, andmeasuredthe recovery angle after twisting.
For a wire of length 243.6mm and diameter 0.51mm, the shearing
elastic modulus was estimated to be about 8200kgf/mm
2
.
Fig. 2 shows the relation between the number of rotations in
twisting and the angle of spring back. When the number of rota-
0924-0136/$ see front matter 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmatprotec.2010.04.001
K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454 1437
Fig. 1. Torsion test by Coulomb (Bell, 1984).
tions exceeds about 0.5, the angle of recovery becomes smaller
than the angle of twisting, and the recovery angle increases only
slightly when the number of rotation exceeds two times. This phe-
nomenon suggests that plastic deformation starts on the surface
of the wire when the rotation is about 0.5, and then the plastic
zone expands towards the centre of the wire up to two rotations,
and work-hardening proceeds gradually as the number of rotation
increases further.
Let us estimate the yield stress and the ow stress of this wire
from its dimensions and the elastic modulus. The shear strain and
stress after 0.5 rotations are calculated respectively to be 0.003 and
24kgf/mm
2
, whichis a little larger thanthe presentlyknownshear-
ing yieldstress of iron, but may be reasonable if plastic deformation
has already proceeded for 0.5 rotations. For the spring back angle
of 450

, the shear stress is calculated to be k =50kgf/mm

2
if it is
distributed uniformly across the cross-section. This shearing ow
stress seems to be reasonable, too.
Fig. 2. Number of rotation and recovery angle in twisting of iron wire carried out
by Coulomb.
Fig. 3. C.A. Coulomb.
Charles A. de Coulomb (17361808) (Fig. 3) entered the military
corps of engineers after receiving preliminary education in Paris.
He was sent to the island of Martinique in the West Indies for 9
years. There he studied the mechanical properties of materials. In
1773, he submitted his rst paper on the fracture of sandstone to
the Academy. He concluded that fracture of sandstone occurred
when the shear stress reached a certain value, similarly to the yield
condition due to maximum shear stress.
After returning to France, he worked as an engineer, and con-
tinued to carry out research. In 1781, he won an Academy prize for
his paper on friction, now known as Coulomb friction, and in the
same year he was elected to membership of Academy.
2.3. Elasticity and stressstrain curve
In the early 19th century, the mathematical theory of elastic-
ity began to ourish due to efforts of the scholars related with
the cole Polytechnique such as S.D. Poisson (17811840), Navier
(17851836), A. Cauchy (17891857) and G. Lme (17951870), in
parallel withthose at the University of Cambridge University, being
T. Young (17731829) and G. Green (17931841) (as reported by
Timoshenko, 1953).
To determine the elastic constants, measurements of the
stressstrain relations of metals were begun, and after extension
in the elastic range, stressstrain curves in the plastic range were
measured. Fig. 4 is the stressstrain curve of piano wire measured
by F.J. Gerstner (17561832) and published in 1831 (Bell, 1984). He
applied the load to a piano wire of 0.63mmin diameter and 1.47m
inlengthwitha series of weights. It is obvious that the plastic strain
is measured after unloading.
2.4. H. Tresca (Bell, 1984)
H.E. Tresca carried out experiments on metal forming such as
punching, extrusion and compression using various metals, and
measured the relation between the forming load and ramdisplace-
ment. He presented a series of papers to the French Academy of
Sciences, starting in1864. In Fig. 5, the cross-section of an extruded
1438 K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454
Fig. 4. Stressstrain curve of piano wire measured by Gerstner (Bell, 1984).
billed made of 20 lead sheets is given. Tresca was interested in the
metal ow as suggested by the title of his rst paper, Mmoire sur
lncoulement des corps solides de fortes pression (On the ow
of a solid body subjected to high pressure), rather than yielding in
material testing.
Tresca assumed that the extrusion force P could be expressed in
terms of the shear stress k, and estimated the value of k from the
measured forming load of various processes. Because the values
of shear stress k estimated from the forming loads occurred in a
certain range, he concluded that the metal ow occurred under a
constant maximum shear stress. The values of shearing ow stress
measured by Tresca are given in Table 1. It seems that the ow
stress values are reasonable even from the present view point.
Henri E. Tresca (18141885) (Fig. 6) graduated from the cole
Polytechnique at the age of 19 in 1833, and sought a career in the
design of civil structures. But his ambition was deterred by serious
illness, and he spent many years teaching, building and performing
tests on hydraulics. In 1852 he began to work at Conservatoire des
Arts et Metie in Paris as an engineer. He suddenly started research
work when he was promoted to a major experimental physicist
at the age of 50, and soon published many papers. After 8 years
Fig. 5. Extruded rod by Tresca (Bell, 1984).
Table 1
Shearing ow stress measured by Tresca.
Material Shearing ow stress (kgf/mm
2
)
Lead 1.82
Pure titanium 2.09
Leadtitanium alloy 3.39
Zinc 9.00
Copper 18.93
Iron 37.53
concentrated research activity, he was elected as a member of the
French Academy of Sciences.
2.5. Saint-Venant and Lvy (Timoshenko, 1953)
When Tresca presented his paper to the French Academy of Sci-
ences, Barr de Saint-Venant (17971886) was the authority of
mechanics in France, elected a member of the Academy in 1868.
After reading the experimental results of plastic owby Tresca, his
attention was drawn to the area of plasticity. In 1871, he wrote a
paper on elasticplastic analysis of partly plastic problems, such as
the twisting of rods, bending of rectangular beams andpressurizing
of hollow cylinders.
Saint-Venant assumed that (1) the volume of material does not
change during plastic deformation, (2) the directions of principal
strains coincide with those of the principal stresses (now known
as total strain theory), and (3) the maximum shear stress at each
point is equal to a specic constant in the plastic region.
The last assumption is now known as the Tresca yield crite-
rion which is expressed with the maximumprincipal stress
1
, the
minimum principal stresses
3
and ow stress Y as:

3
= 2k = Y (1)
Although his analyses are not complete from our current point
of view, it can be said that plastic analysis started from this paper.
Fig. 6. H.E. Tresca.
K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454 1439
Fig. 7. Bauschinger effect.
2.6. Bauschinger and Mohr (Timoshenko, 1953)
In the second half of the19th century, the Technical Universi-
ties in German speaking areas became important research centres
in plasticity and metal forming. They were established as PS:
Polytechnische Schule and then changed to university level TH:
Technische Hochschule.
Johann Bauschinger (18331893) graduated from Munich Uni-
versity and became a professor of Munich PS in 1868. He installed
a 100tons tensioncompression universal testing machine with an
extensometer of his own invention, and carried out a vast num-
ber of measurements of stressstrain relations. He found that the
yield stress in compression after plastic tensile deformation was
signicantly lower than the initial yield stress in tension. Fig. 7 is
the experimental result from tests carried out in 1885, in which
the compression test is performed after a tension test up to a strain
of 0.6%. It is seen that the initial yield stress was 20.91kgf/mm
2
and the yield stress in compression after tensile deformation was
9.84kgf/mm
2
.
In 1882, Otto Mohr (18351913) presented a graphical repre-
sentationof stress at apoint. Onagraphwithaxes indicatingnormal
and shear stress components, the stress state of a point on a plane
is expressed by a circle. Mohr used his representation of stress to
devise a strength theory.
Fig. 8 shows the stress circles for cast iron tested in tension,
compression, and in torsion. Mohr suggested that the envelope of
the circles was a fracture limit. This idea was extended to a yield
conditioninwhichshearing yieldstress was affectedby hydrostatic
pressure. This condition is often called Mohrs yield condition.
Mohr graduated from Hannover PS and worked as a structural
engineer. When he was 32 years old, he was already a well-known
engineer and was invited by Stuttgart TH. After teaching engi-
neering mechanics there until 1873, he moved to Dresden TH and
continued teaching.
Fig. 8. Mohrs stress circle.
Fig. 9. Experimental result plotted on principal stress plane by Guest (1900).
2.7. J. Guest (Guest, 1900)
In 1900, James Guest (University College London) published a
paper from the Royal Society on the strength of ductile materials
under combined stress states. By carrying out tension and torsion
tests of internally pressurizedtubes, he examinedthe occurrence of
yielding. Guest is the rst person to differentiate yielding of ductile
metal frombrittlefracture, wherepreviouslyfailure hadbeenused
to express the strength limit of material both due to yielding and
brittle fracture.
He came to the conclusion that yielding occurs when the max-
imum shear stress reaches a certain value. In Fig. 9, the yielding
points are plotted on a graph of principal stresses in plane stress.
Although his conclusion was the same as that of H. Tresca, he natu-
rally thought that he had found a newcriterion of yielding because
he did not recognize that the large plastic ow observed by Tresca
and the initial yielding he observed were essentially the same phe-
nomenon.
3. Yield criteria and constitutive equations
3.1. Progress of research in yield condition
During the 19th century, the maximum shear stress criterion
was established by Tresca, Saint-Venant, Mohr and Guest. The yield
criterion of elastic shear-strain energy, mostly called Mises yield
criterion, was put forward in the early 20th century. It is written by
the following equation with the maximum, mediumand minimum
principal stresses
1

3
as:
1

2
_
(
1

2
)
2
+(
2

3
)
2
+(
3

1
)
2
= Y (2)
In 1904, M.T. Huber proposed this criterion (Engel, 1994)
although limited to compressive hydrostatic stress conditions. This
paper was not known for 20 years by the researchers of plasticity
1440 K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454
because it was written in the Polish language. von Mises (1913)
wrote his paper from the view point of mathematics without dis-
cussing the physical background. Hencky (1924) introducedHubes
paper and derived the yield criterion of elastic shear-strain energy.
Ndai (1937) showed that the criterion could be interpreted as a
prediction that yielding occurs when the shear stress on the octa-
hedral plane in the space of principal stresses reaches a critical
value. This idea is nowoftenused intext books of plasticity because
of its simple graphical representation, although it has no physical
meaning.
The difference between the Tresca yield criterion Eq. (1) and
the Mises criterion Eq. (2) when expressed with principal stresses
is that the medium principal stress (
2
) has an effect (Mises) or
does not (Tresca). This difference was experimentally examined by
Lode (1926) following a suggestion by Ndai, and then Taylor and
Quinney (1932). Their conclusions were that the mediumprincipal
stress didinuence the yieldingconditionfor mildsteel, aluminium
and copper and thus the Mises criterion offered a better approxi-
mation for the yield condition.
Hill (1948) proposed a yield criterion for anisotropic materi-
als, and since then many researchers have also tried to express
the yielding behaviour of anisotropic materials. Yield criteria for
other materials which are not incompressible and isotropic have
also been put forward: one example is the criterion for porous or
compressible metals proposed by Shima and Oyane (1976).
3.2. Constitutive equation
In his paper published in 1872, M.Lvy (according to
Timoshenko, 1953) used an incremental constitutive equation. von
Mises proposed the same constitutive equation because Lvys
paper was not known outside of France. Mises considered that the
increments of plastic strain components d
p
1
, d
p
2
, d
p
3
were in pro-
portion to the deviatoric stress components

1
,

2
,

3
, where for
example

1
=
1
(
1
+
2
+
3
) /3. Thus,
d
p
1

1
=
d
p
2

2
=
d
p
3

3
(3)
In the plastic strain range of an elasticplastic material, the
increments of elastic strain components d
e
1
, d
e
2
, d
e
3
, and the plas-
tic strain increments d
p
1
, d
p
2
, d
p
3
, should be handled separately.
Prandtl (1924) treated this problem for plane-strain, and Reuss
(1930) (Budapest Technical University) showed the expression for
all of the strain components. For example,
d
1
= d
e
1
+d
p
1
=
1
E
_
d
1
(d
2
+d
3
)
_
+
d
Y
_

1
2
(
2
+
3
)
_
(4)
where d is an equivalent strain increment which is expressed
in terms of the plastic strain increments, and Y is the ow stress.
In the 1960s, when the nite element analysis of elasticplastic
material was under development, a key topic was the inversion of
the above PrandtlReuss equation to express the stress increments
in terms of strain increments. However it was eventually found
that R. Hill had already done this work in his book published in
1950 (Hill, 1950).
3.3. Letter of J. Maxwell (Timoshenko, 1953)
The letters from James Clerk Maxwell (18311879: famous for
Maxwells equation) to his friend William Thomson (Lord Kelvin:
18241907) were published in 1937 in a book, and it was then
found that Maxwell had written about the occurrence of yielding
as early as 1856.
Maxwell showed that the total strain energy per unit volume
could be resolved into two parts (1) the strain energy of uniform
tension or compression and (2) the strain energy of distortion. The
total elastic energy per unit volume is expressed as:
F =
1
2
E
3(1 2)
(
1
+
2
+
3
)
2
+
1
3
E
2(1 +2)
_
(
1

2
)
2
+(
2

3
)
2
+(
3

1
)
2
_
(5)
where E is Youngs modulus, is Poissons ration and,
1
,
2
,

3
are principal stresses. The rst term on the right side of the
equation is the energy for volume change due to uniform tension
or compression, and the second term is the energy of distortion.
Maxwell made the statement: I have strong reasons for believ-
ing that when the strain energy of distortion reaches a certain limit,
then the element will begin to give way. Further on the states:
This is the rst time that I have put pen to paper on this sub-
ject. I have never seen any investigation of the question, given the
mechanical strain in three directions on an element, when will it
give way? Unfortunately he did not return to this subject again.
3.4. M.T. Huber (Engel, 1994; Olesiak, 2000)
In 1904, M.T. Huber proposed that yielding was determined by
elastic shear-strain energy distortion when the hydrostatic stress
was compressive, and by the total elastic energy when the hydro-
static stress was tensile. Hubers paper in the Polish language did
not attract general attention until H. Hencky introduced it in 1924.
MaksymilianTytus Huber (18721950) (Fig. 10) graduatedfrom
Lww (now Lviv, Ukraine) Technical University in 1895, and stud-
Fig. 10. M.T. Huber.
K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454 1441
ied mathematics at Berlin University. In 1899 he began to work
at a technical school in Krakw and wrote his paper on the yield
condition. In 1909 he was invited to head the Chair of Technical
Mechanics at Lww Technical University.
When World War I began, he was called to the AustriaHungary
army but captured by Russian troops. After World War I, he went
back to LwwTechnical University and became the president of the
university. In 1928 he moved to Warsaw Technical University, and
actively took part invarious advisory and expert bodies. He became
a member of the Polish Academy of Learning.
During World War II, he was not able to work in the university
and lost all his belongings in a re ignited by Germans troops dur-
ing the Warsaw uprising of 1944, but after the war he was able to
continue research work in Gdansk Technical University.
3.5. R. von Mises
von Mises (1913) manipulated the maximum shear stresses on
the principal stress planes:

1
=

3

2
2
,
2
=

1

3
2
,
3
=

2

1
2
(6)
It is apparent that simple summation of the maximum shear
stresses is always zero:

1
+
2
+
3
= 0 (7)
In the space of the maximum shear stresses, he expressed the
criterion of maximum shear stress:
|
1
| k, |
2
| k, |
3
| k (8)
which Mises called Mohrs yield criterion, and is represented in
the cube shown in Fig. 11. The yield condition of maximum shear
stress is expressedas the intersectionof the cube andEq. (7) is given
by the hexagon in the gure.
Then he considered a sumof squares of the shear stresses as the
sphere in the gure.
k
2
=
2
1
+
2
2
+
2
3
=
1
4
_
(
1

2
)
2
+(
2

3
)
2
+(
3

1
)
2
_
(9)
The circle resulted as the intersection of the sphere and Eq. (7)
is an approximation of the Mohr (Tresca) criterion. While the Mohr
criterion cannot be expressed by a simple mathematical equation,
the newcriterion is easy to handle mathematically, as is often done
in mathematical plasticity.
Richard von Mises (18831953) (Fig. 12) was born in Lemberg
(nowLviv, Ukraine) andgraduatedinmathematics fromthe Vienna
University of Technology. In 1908 Mises was awarded his doctorate
from Vienna. In 1909, at the age of 26, he was appointed professor
in Straburg (now Strasbourg, France) and received Prussian cit-
izenship. There he wrote his paper on the yield criterion. During
World War I, he joined the AustroHungarian army and ew as a
Fig. 11. Handling of maximum shear stresses by Mises.
Fig. 12. R. von Mises.
test pilot, and then supervised the construction of a 600HP aircraft
for the Austrian army.
After the war Mises held the new chair at Dresden TH. In 1919,
he was appointed director of the new Institute of Applied Mathe-
matics created in the University of Berlin. In 1921 he became the
editor of the newly founded journal Zeitschrift fr Angewandte
Mathematik und Mechanik and stayed until 1933.
With the rise of the Nazi party to power in 1933, Mises felt his
position threatened. He moved to Turkey, where he held the newly
created chair of Pure and Applied Mathematics at the University of
Istanbul. In 1939, amid political uncertainty in Turkey, he went to
the USA, where he was appointed in 1944 as the Gordon-McKay
Professor of Aerodynamics and Applied Mathematics at Harvard
University.
3.6. A.L. Ndai
Arpad L. Ndai (18831963) (Fig. 13) was born in Hungary and
graduated fromBudapest University of Technology, and then stud-
ied in Berlin TH getting his doctorate in 1911. In 1918 he moved
to L. Prandtls Institute of Applied Mechanics in Gttingen and was
promoted to professor in 1923. In 1927, he moved to the Westing-
house Laboratory in the USA as the successor of P.E. Timoshenko.
Thus his paper in 1937 on the yield criterion was based on work
performed in the USA.
In 1927, Ndai published a book of plasticity in German and
this was translated into English as Plasticity A Mechanics of the
Plastic State of Matter (Ndai, 1931) as the rst English book of
plasticity. The characteristic feature of this book is that it consists
of two parts, (1) plasticity of metals and (2) application of plastic-
ity in geophysics problems. In 1950 the rst part of this book was
rewritten and published as Theory of Flowand Fracture of Solids.
1442 K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454
Fig. 13. A.L. Ndai.
4. Physics and metallurgy of plastic deformation
4.1. Plastic deformation of a single crystal
In 1923, P.W. Bridgman invented a method to make a single
crystal of metal by pulling it out of molten metal. Since M. von Laue
(18791960) had already established the method for determining
the direction of the crystal lattice by X-ray diffraction, the study
of plastic deformation of single crystals could immediately begin.
Taylor and Elam (1923) carried out a tension test of an Al single
crystal (Fig. 14), and found that plastic deformation occurred by
sliding on a certain crystallographic (sliding) plane in a denite
(sliding) direction, and the critical shear stress
cr
on the plane was
calculated. They continuedexperiments withsingle crystals of iron,
gold, copper and brass. In Germany, the groups of Schmid (1926)
and Gler and Sachs (1927) presented their results from tension
tests of single crystals.
Sachs (1928) calculated the yield stress of polycrystalline metal
as an average of those of single crystals with random orienta-
tions. The calculated average yield stresses in tension and shearing
were
y
= 2.24
cr
and
y
= 1.29
cr
, respectively, and the ratio was

y
/
cr
= 0.577. This ratio was the same as that derived from Mises
yield condition, but the calculated constant 2.24 was too small
compared with the experimental value. Taylor (1938) proposed a
Fig. 14. Stressstrain curve of single crystal by Taylor and Elam (Bell, 1984).
Fig. 15. Rotation of crystal by plastic sliding.
method to relate the yield stress of polycrystalline metals with that
of single crystals by taking account of the constraints provided by
neighbouring grains, giving
y
= 3.96
cr
, which was quite near to
the value obtained by experiments.
When a single crystal is plastically stretched, the direction of
the crystal rotates as demonstrated in Fig. 15 due to sliding over
the specic planes, and the sliding planes tend to become parallel
to the stretching direction irrespective of the initial orientation.
This means that anisotropy is developed by plastic deformation of
polycrystalline metals. Boas and Schmid (1930) were the rst to
study the development of anisotropy.
4.2. Dislocation theory
When the initially polished surfaces of single crystals were
observed after plastic deformation, slip bands (Fig. 16) were
observed suggesting that sliding occurred on a limited number of
sliding planes. Since an extremely large shear stress, 100010,000
times as large as the measured critical shear stress, would be
needed to overcome the atomic bonding stress, many researchers
worked to understand the mechanism of plastic deformation.
Taylor (1934), Polanyi (1934), and Orowan (1934) independently
proposedthe sliding mechanismby crystal defects, i.e. dislocations.
Fig. 17shows the explanationby Taylor about dislocations ina crys-
tal lattice during plastic deformation. The existence of dislocations
was proved in the 1950s after electronic microscopy was invented.
Whenthegeneral assemblyof theInternational Unionof Physics
was held in Tokyo in 1953, N.F. Mott (19051999), the president
of the Union, told the delegates that the rst person to recognize
the existence of dislocations was K. Yamaguchi. Yamaguchi (1929)
showed a representation of dislocations as shown in Fig. 18 to
explain the cause of the warping of a single crystal after plastic
deformation. Yamaguchi carried out the research in the Institute of
Physics and Chemistry at the laboratory of M. Mashima, who had a
close relation with the laboratory of G. Sachs in Germany. In 1937,
Yamaguchi was appointeda professor of Osaka Imperial University,
when it was established.
K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454 1443
Fig. 16. Slip bands on polished single crystal.
4.3. Response of metals to high deforming speed
The measurement of stressstrain curves began to attract
researchers in the second half of the 19th century but appropriate
measuring methods for high speed phenomena did not exist. Dunn
(1897) carried out compression tests by using a drop hammer, and
measured the displacement of the hammer optically and recorded
it on a lm attached to a rotating drum. By differentiating the dis-
Fig. 17. Dislocation model proposed by Taylor and Elam (1923).
Fig. 18. Dislocation model suggested by Yamaguchi (1929).
placement, he calculated the hammer velocity, and then obtained
the acceleration or force applied to the hammer by differentiating
the velocity. From the measured displacement and the calculated
force, he was able to determine the stressstrain curves. In the
early 20th century, high speed stressstrain curves were obtained
by some groups in Europe with similar measuring methods.
Itihara (1933) (Tohoku Imperial University) measured the
shearing stressshearing strain curve at high speed and high tem-
perature up to 1000

C. Fig. 19 shows the equipment of the torsion

test in which the torque was determined by the twisting angle of
the measuring bar.
Manjoine and Ndai (1940) measured the stressstain curves in
a high speed tension test at up to 1000

C as shown in Fig. 20 by
using a load cell with strain gauges.
Kolsky (1949) used the split Hopkinson bar, which was devel-
oped by B. Hopkinson at the University of Cambridge in 1914 (Bell,
1984). To measure a high strain rate stressstrain curve, a spec-
imen was sandwiched between two long bars, one end of which
was struck and the transmitted elastic wave was measured at the
other.
4.4. P.W. Bridgman (Bridgman, 1964)
Although von Krmn carried out compression tests on mar-
ble under high pressure and published the results in 1911, the
mechanical behaviour of metals under high hydrostatic pressure
was mainly studied by P.W. Bridgman during the rst half of the
20th century. Although he found that the ductility of metal was
Fig. 19. High speed torsion test used by Itihara (1933).
1444 K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454
Fig. 20. Stressstrain curves at high speed tension test by Manjoine (1940).
Fig. 21. Fracture strain and pressure measured by Bridgman (1964).
remarkably enhanced by pressure as Fig. 21, he was more inter-
ested in the effect of pressure on the stress, which is only a little
affected by pressure as shown in Fig. 22.
Percy Williams Bridgman (18821961) (Fig. 23) studied physics
in Harvard University and received his Ph.D. in 1908. He was
appointed Instructor (1910), Assistant Professor (1919), before
becoming Hollis Professor of Mathematics and Natural Philoso-
phy in 1926. He was appointed Higgins University Professor in
1950. From1905, Bridgman continued the experiments under high
pressure for about 50 years, and published the results on plastic
deformation of metals in the book Studies in Large Plastic Flow
andFracture. He inventeda methodfor growingsingle crystals and
proposed a calculation method for the stress state in the neck of a
tensile test specimen. A machinery malfunction led him to modify
his pressure apparatus; the result was a new device enabling him
to create pressures eventually exceeding 100,000kgf/cm
2
(10GPa).
This newapparatus brought about a plenty of newndings, includ-
ing the effect of pressure on electrical resistance, and on the liquid
and solid states. In 1946, he received the Nobel Prize in physics for
his work on high pressure physics.
Fig. 22. Average stress in tension test as a function of pressure measured by Bridg-
man.
Fig. 23. P.W. Bridgman.
4.5. G.I. Taylor
Geoffrey Ingram Taylor (18861975) (Fig. 24) was born in Lon-
don, and studied mathematics at Trinity College, Cambridge. At the
outbreak of World War I he was sent to the Royal Aircraft Factory
at Farnborough to apply his knowledge to aircraft design.
Fig. 24. G.I. Taylor.
K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454 1445
After the war, Taylor returned to Trinity working on an applica-
tion of turbulent ow to oceanography. In 1923, he was appointed
to a Royal Society research professorship as a YarrowResearch Pro-
fessor. This enabled him to stop teaching which he had been doing
for the previous 4years andwhichhe bothdislikedandfor whichhe
hadno great aptitude. It was inthis periodthat he didhis most wide
ranging work on the deformation of crystalline materials which led
on from his war work at Farnborough.
During World War II Taylor again worked on applications of
his expertise to military problems. Taylor was sent to the United
States as part of the British delegation to the Manhattan project.
Taylor continued his research after the end of the War serving on
the Aeronautical Research Committee and working on the devel-
opment of supersonic aircraft. Though technically retiring in 1952,
he continued researching for the next 20 years.
4.6. M. Polanyi
Michael Polanyi (18911976) was born into a Jewish family in
Budapest, Hungary and graduated frommedical school of Budapest
University. His scientic interests led himto further study in chem-
istry at the Karlsruhe TH in Germany and he was awarded his
doctorate in 1917. In 1920, he was appointed a member of the
Kaiser Wilhelm Institute for Fibre Chemistry, Berlin, where he
developed new methods of X-ray analysis and he made contribu-
tions to crystallography including dislocation theory.
In 1933, he resigned his position in Germany when Hitler came
to power. Within a few months he was invited to take the chair of
physical chemistry at the University of Manchester in England. He
believed from his experience in science that there was a necessary
connectionbetweenthepremises of afreesocietyandthediscovery
of scientic truths. In 1938 he formed the Society for the Freedom
of Science.
4.7. E. Orowan
Egon Orowan (19021989) (Fig. 25) was born in Budapest and
received his doctorate from Berlin TH on the fracture of mica in
1932. He had difculty in nding employment and spent the next
few years ruminating on his doctoral research, and completed his
paper on dislocations.
After working for a short while on the extraction of krypton
fromthe air for the manufacture of light bulbs in Hungary, Orowan
moved in 1937 to the University of Birmingham where he worked
on the theory of fatigue collaborating with R. Peierls (19071955).
In 1939, he moved to the University of Cambridge where W.L.
Bragg (18901971: X-ray analysis) inspired his interest in X-
ray diffraction. During World War II, he worked on problems in
munitions production, particularly developing an understanding
of plastic ow during rolling. In 1950, he moved to MIT where,
in addition to continuing his metallurgical work, he developed his
interests in geological and glaciological deformation and fracture.
5. Slip-line eld method
5.1. Progress of the slip-line eld method
Prandtl (1920) presented an analysis of the plane-strain inden-
tation of a at punch into a rigidplastic solid body as shown in
Fig. 26. He assumeda rigid-perfectly plastic material without work-
hardening but with a pressure sensitive ow stress (Mohr yield
criterion). By solving the equilibrium equation, he constructed a
series of lines having directions parallel to the maximum shear
stress as showninFig. 26. He correctly obtained the indenting pres-
sure for a material with shearing ow stress k without pressure
Fig. 25. E. Orowan.
sensitivity as:
p = 2k
_
1 +

2
_
(10)
Hencky (1923) derived a general theorem of stress states for
slip-line elds which now carries his name. A statically admissible
stress eldwhichsatises theequilibriumequation, yieldcondition
and boundary force is not always correct, because the velocity eld
associatedwiththestress statemaynot satisfyvolumeconstancyor
mayleadtonegativeenergyconsumption. Geiringer (1930) derived
an equation in relation to the velocity eld by considering the
incompressibility condition in plastic deformation and the relation
between strain rate and velocity.
In 1933, when the fundamentals of slip-line eld theory were
established, the Nazis came to power in Germany and forced the
Jewish researchers to leave from university positions. The remark-
able progress attained in the eld of plasticity was thus halted in
Germany. The researchers expelled from Germany tried to nd a
safehaveninTurkey, theUnitedStates andEngland, whereresearch
work on plasticity was replanted.
During World War II, R. Hill used the slip-line eld method to
predict the plastic deformation of a thick plate being penetrated by
a bullet. He proposed slip-line elds for various problems such as
wedge indentation, compression of thin plates with friction, plate
drawing (Fig. 27) andtensiontests for a notchedplate. He usedslip-
Fig. 26. Slip-line eld by Prandtl (1920).
1446 K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454
Fig. 27. Slip-line elds for drawing and extrusion proposed by Hill (1950).
line theory which had been developed as a mathematical method
for the purpose of practical engineering purposes.
Sokolovskii (1948) reported that active research works were
performed in the area of slip-line elds in the Soviet Union too. This
is possibly due to the inuence of H. Hencky, who established the
slip-line theory and stayed in the Soviet Union to carry out research
until 1938.
Hill (1950), and Prager and Hodge (1951), rst presented a sys-
tematic account of slip-line theory and displayed the engineering
worth of the approach. Pragers introduction of the hodograph, or
velocity plane diagram, in 1953 introduced a vast simplication
into the handling of slip-line solutions andremovedthe difculties.
During the 1950s and 60s, many new slip-line elds were pro-
posed (Johnson et al., 1970) for extrusion, rolling, drawing and
metal cutting. Since the slip-line method was the only mean to
allow prediction of the stress state in the deforming material at
that time, it was used widely even though the required assumption
of plane-strain behaviour was not realistic for most metal forming
operations. When nite element methods enabled precise stress
calculation in axi-symmetric and later 3D problems, the use of the
slip-line eld method decreased from around 1980, although its
academic value was not lost.
5.2. L. Prandtl
Ludwig Prandtl (18701953) (Fig. 28) received his engineer-
ing education at the Munich TH. After graduating, he remained
at the school as an assistant of A. Fppl (18541924: successor
of J. Bauschinger), and carried out doctoral work on the bending
of circular plates. After working in industry for a while, he was
appointed as a professor of industrial mechanics at Hannover THin
1900. There he proposed the membrane analogy of torsion and the
boundary layer of uid ow. In 1904 he was invited to the Insti-
tute of Mechanics in Gttingen University. Soon he began to study
plasticity such as plastic buckling and bending. He was appointed
the leader of the laboratory of aerodynamics, and studied wing
theorems and other important works of uid dynamics.
In 1922, Prandtl established the society of applied mathemat-
ics and mechanics, Gesellschaft fr Angewandte Mathematik und
Mechanik, andledthe area of appliedmechanics. He is alsofamous
as the teacher of many leaders in mechanics in the 20th century
such as Th. von Krmn (California Institute of Technology), S.P.
Timoshenko (Stanford University), A. Ndai (Westinghouse Labo-
ratory), W. Prager (Brown University) and others.
5.3. H. Hencky
Heinrich Hencky (18851951) (Fig. 29) graduated from Darm-
stadt TH and began to work in Ukraine as an engineer of a railway
company in 1913 at the age of 28. Soon World War I began and the
area was occupied by Russian troops, and he was kept in a camp in
Ural, where he married a Russian woman.
Although he could not nd a permanent job after the war in
Germany, he was awarded his Habilitation (qualication for pro-
fessorship) from Dresden TH and found a job in Delft Technical
Universityin1922. Hecarriedout researchintoslip-lineeldtheory
in Delft, and stayed until 1929.
In 1930, he moved to MIT in the USA, but his scientic approach
to engineering was not accepted there because practical tech-
nologies were overwhelming, and he resigned from MIT after
only 2 years. In 1936, Hencky was invited to the Soviet Union
by B.G. Galerkin (18711945: variational method) and carried out
research in Moscow University. But in 1938, as relations between
the Soviet Union and Germany worsened, he returned to Germany,
and worked in a bus manufacturing company in Mainz.
Fig. 28. L. Prandtl.
K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454 1447
Fig. 29. H. Hencky.
5.4. H. Geiringer
Hilda Geiringer (18931973) (Fig. 30) was born in Vienna and
received her doctorate in 1917 from the University of Vienna with
a thesis about Fourier series. From 1921 to 1927 she worked at
the Institute of Applied Mathematics in the University of Berlin
Fig. 30. H. Geiringer.
as an assistant of von Mises. Her mathematical interests had
switchedfrompure mathematics to probability andthe mathemat-
ical development of plasticity theory. In 1927, Geiringer became
Privatdozent (lecturer). During this period she had a brief marriage
and had one daughter.
In 1930, her work in plasticity theory led to the development of
the fundamental Geiringer equations for plane-strain plastic defor-
mations. Geiringer remained at the University of Berlin until forced
to leave when Hitler came to power. After a brief stay as a research
associate at the Institute of Mechanics in Belgium, she became a
professor of mathematics at Istanbul University in Turkey where
she stayed for 5 years.
In 1939, she emigrated to the United States with the help of
A. Einstein, and became a lecturer at Bryn Mawr College. While
at Bryn Mawr she married R. von Mises who was then teaching
at Harvard. In 1944, Geiringer became professor and chair of the
mathematics department at Wheaton College in Massachusetts.
Attempts to nd a position at some of the larger universities near
Boston repeatedly failed, often because of her gender. From1955 to
1959, she worked as a research fellow in mathematics at Harvard
in addition to her position at Wheaton to complete her husbands
unpublished manuscripts Mathematical Theory of Probability and
Statistics after his death in 1953. Geiringer was elected a fellowof
the American Academy of Arts and Science.
5.5. W. Prager (Hopkins, 1980)
William Prager (19031980) (Fig. 31) was born in Karlsruhe,
and studied at Darmstadt TH receiving his doctorate in 1926 at
the age of 23. From 1929 to 1933 he worked as the acting director
of Prandtls Applied Mechanics Institute at Gttingen. At the age of
29, he was appointed at Karlsruhe TH as the youngest professor in
Germany, but soon he was dismissed when Hitler came to power.
He was invited to Istanbul University, Turkey and acted as a special
adviser in education to the government. Prager remained in Istan-
bul until 1941. The expansion of World War II made the position
Fig. 31. W. Prager.
1448 K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454
Fig. 32. R. Hill.
of the German refugees insecure, and he accepted the invitation of
Brown University in the USA made on the recommendation of A.
Einstein.
The Graduate Divisionof AppliedMathematics at BrownUniver-
sity was created in1946 withPrager as its rst Chairman, a position
he held until 1953. By his effort, Brown University became the cen-
tre of applied mechanics, especially in the area of plasticity in the
1950s and 60s.
5.6. R. Hill (Hopkins and Sewell, 1982)
Rodney Hill (1921) (Fig. 32) was born in Yorkshire, England
and read mathematics at Gonville and Caius College, Cambridge,
where E. Orowan was teaching. In 1943, amid World War II, he
joined a theoretical group on armaments led by N. Mott and he was
assigned the problemof deep penetration of thick armour by high-
velocity shells. This aroused Hills interest in the eld of plasticity.
From 1946, he began to work with the group of metal physicists
under E. Orowan at the Cavendish Laboratory in Cambridge. He
solved various metal forming problems using plasticity theory, and
obtained his Ph.D. in 1948. In 1949, he was invited to be the head of
a new Section of the Metal Flow Research Laboratory of the British
Iron and Steel Association (BISRA).
Hill expanded his Ph.D. thesis and published the book The
Mathematical Theory of Plasticity in 1950 when he was 29 years
old. This bookwas soonacceptedas a standardof mechanics of plas-
ticity. In 1952, he became the Editor in Chief of a new journal, the
Journal of Mechanics and Physics of Solids, which was eventually
knownas the highest level journal inmechanics. In1953 he applied
for and was offered the post of a newChair of Applied Mathematics
in Nottingham University, and undertook administrative work on
top of the research works of plasticity until his retirement fromthe
university in 1962. From 1963, Hill moved back to Cambridge and
continued research work in solid mechanics.
Fig. 33. Model of slab method used by Siebel (1923).
6. Slab method
6.1. Analysis of forging by E. Siebel
Siebel (1923) wrote a paper on the analysis of forging. He
assumed a thin area (slab) on which to dene an equilibriumequa-
tion. In the case of the compression of a cylinder of diameter d and
height 2h shown in Fig. 33, he considered a thin layer of a thickness
dx anda height the same as the cylinder, anddenedanequilibrium
of forces acting on the layer.
For the case of the yield stress Y and friction coefcient , he
derived the average contacting pressure

P as:

P = Y
_
1 +
1
3

_
d
h
__
(11)
Using more recent applications of the slab method, the average
pressure is calculated to be

P = 2Y
_
h
d
_2 _
exp
_
d
2h
_

d
2h
1
_

= Y
_
1 +
1
3

_
d
h
__
(12)
The last equation, identical to Siebels solution, is anapproxima-
tion of the second equation when d/h is sufciently smaller than
1.0. Siebel numerically calculated the average pressures for some
typical cases of forging, and discussed ways to apply the result to
backward extrusion.
Soon after Siebels paper, similar methods were used by von
Krmn (1925) for analyzing rolling of sheet metal and by Sachs
(1927) for solving wire drawing. Using this method, Siebel contin-
ued to analyze various processes, and many researchers extended
the method. Since the results are mathematically analytical, they
are widely used in industry (Lippmann and Mahrenholts, 1967;
Lange, 1985).
Erich Siebel (18911961) (Fig. 34) received his doctorate from
Berlin TH in 1923 at the age of 32, on the topic of the calculation
of load and energy in forging and rolling. After working in the steel
industry for a short time, he became a leader of the metal forming
division at Kaiser Institute in Dusseldorf (steel), and carried out
analysis of rolling and forging. In 1931, he became a professor of
Stuttgart TH, and treated almost all areas of metal forming such as
deep drawing and wire drawing. He continued theoretical research
into metal forming until retirement in 1957.
K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454 1449
Fig. 34. E. Siebel.
6.2. Th. von Krmn (von Krmn, 1967)
In 1925, Th. von Krmn presented a short paper on rolling to
a meeting of applied mechanics. In three pages, the fundamental
differential equation, a result of pressure distribution and energy
efciency were given. Although this paper gave a great inuence
to the subsequent researches in rolling technology, Krmn never
returned to this subject again.
Theodore von Krmn (18811963) (Fig. 35) was born into
a Jewish family in Budapest, Austria Hungary and he studied
engineering at the Royal Technical University in Budapest. After
graduating in 1902, he joined Prandtls Institute at Gttingen Uni-
versity, and received his doctorate in 1908. Then he taught at
Gttingen for 4 years. During this period he measured stressstrain
curves for marble under high pressures. He was also interested in
vibrationinducedbyuidowandfoundtheKrmnvortex, which
made him famous in the area of uid dynamics.
In 1912, he accepted a position as director of the Aeronau-
tical Institute at RWTH Aachen. His time at RWTH Aachen was
interrupted by service in the AustroHungarian Army 19151918,
where he designed an early helicopter. In his own biography,
he does not mention his work on rolling analysis in 1925, pos-
sibly because his main interest was building a wind tunnel in
Aachen.
In 1927, Krmn stayed in Japan for a while as an advisor to
an airplane company to build a wind tunnel. In 1930, he accepted
the directorship of the GuggenheimAeronautical Laboratory at the
California Institute of Technology (Caltech) and emigrated to the
United States. He is one of the founders of the Jet Propulsion Labo-
ratory, which is nowmanaged and operated by Caltech. In 1946 he
became the rst chairman of the Scientic Advisory Group which
studied aeronautical technologies for the United States Army Air
Forces. At age 81 von Krmn received the rst National Medal of
Science, bestowed in a White House ceremony by President John F.
Kennedy.
Fig. 35. Th. Von Krmn.
6.3. Development of rolling analysis (Lippmann and Mahrenholts,
1967)
For the case of at rolling shown in Fig. 36, von Krmn (1925)
derived the equilibriumequation for the position x, plate thickness
h and roll angle , as:
d
_
h q
2
_
= p
_
tan tan f
_
dx (13)
where q is the horizontal pressure acting within the plate, p is the
pressure acting on the roll surface, and the friction between the
plate and the roll is =tanf. The minus () sign in the equation is
for the entrance side and the (+) sign is for the exit side.
This equation is valid when the plate thickness is small and the
stress in the thickness direction is almost constant, and further,
friction is described by Coulombs law, as is the case for cold rolling
of thin plate.
Trescas yield condition Eq. (1) is written as:
p
1
q = Y (14)
where p
1
is the vertical pressure (perpendicular to q), which can
be calculated from the roll pressure p and the frictional stress p
as:
p
1
= p ptan (+ for entrance side) (15)
Since the parameters x, h and are related to each other when
the roll radius is given, it is necessary to use only one of them in
solving the differential equation. The stresses p, q and p
1
are not
independent, and one of them should be chosen as the variable.
Thus the equation to be solved may be in the form of
_
p/q/p
1
_
= f
_
x/h/
_
(16)
but the result of this problem cannot be presented explicitly.
1450 K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454
Fig. 36. Model of at rolling by von Krmn (1925).
The distribution of roll pressure in Fig. 37 by Krmn was
obtained numerically, but for designing the rolling mills it is nec-
essary to present the rolling torque, roll force and the energy
explicitly. By introducing various approximations, Trinks (1937),
Ndai (1939) and Hill (1950) made approximate mathematical
expressions leading to nomograms to be used in industry.
In applying theoretical results to cold rolling, roll attening can-
not be neglected because the roll is elastically deformed and the
local radius is changed. The method of treating roll attening by
Fig. 37. Roll pressure calculated by von Krmn (1925).
Trinks and Hitchcock (1935) was often used in the subsequent
analyses of rolling.
In hot rolling, the plate thickness is generally large, and thus
the stress state cannot be assumed uniform in the thickness direc-
tion. Further, the friction in hot rolling is generally high, so sticking
or constant friction stress is considered to be suitable. To include
the stress distribution in the thickness direction and a friction law
other than Coulomb friction, Orowan (1943) proposed a more gen-
eralized differential equation than Krmns equation. Since the
solution of this differential equation cannot be expressed explic-
itly, Orowan and Pascoe (1946), Bland and Ford (1948) and many
others worked to establish approximate mathematical expressions
for rolling force, torque, load and necessary power.
In the 1950s in the USA and UK, and in the 1960s in Japan,
automation of strip rolling in the steel industry began, and many
engineers studied and improved these theories in industry.
7. Upper bound method
7.1. Progress of the upper bound method
The upper boundmethodprovides anapproximate formingload
whichis never lower thanthecorrect value. Becauseof this, theload
calculated by this method is safe in selecting forming machines and
designing tools, and thus this method has been used practically.
From the relation between velocity and strain rate, the associ-
ated strain rate distribution can be determined in the deforming
region. With a kinematically admissible velocity eld, which satis-
es the condition of volume constancy and the velocity boundary
condition, together with the ow stress value of the material, an
energy dissipation rate and a forming load greater than or equal
to the correct values are obtained. This is guaranteed by the limit
theorem for rigid-plastic material.
The upper bound theoremcame to be known when it was intro-
duced in the book by Hill (1950) together with other bounding
theorems. Hill states that Markov wrote a paper about the case of
rigid-perfectly plastic material in 1947 in the Russian language.
Let us consider a simple plane-strain case in which a rigid-
perfectly plastic body with a shearing ow stress k is deforming
by external force T due to a tool moving with velocity v. A kine-
matically admissible velocity eldonly withvelocity discontinuous
lines S

d
with sliding v
*
is assumed, where (
*
) means kinematically
admissible eld. The upper bound theorem states:
Tv

_
S

dS

d
(17)
where the left side is the correct working rate and the right side
is the energy dissipation rate for the plastic deformation along the
velocity discontinuities. This inequality means that the energy dis-
sipation rate of the right side is greater or equal to the correct value
of the left side. The upper bound value of the forming load T is
obtained by dividing the calculated value of the right side by v.
Green (1951) applied this theoremto plane-strain compression
between smooth plates, and compared the result with that of the
slip-line eld method as shown in Fig. 38. While a slip-line eld
requires a long time to draw, a kinematically admissible eld with
only velocity discontinuity lines can be constructed easily with the
help of a hodograph, and gives a reasonably good result.
With the upper bound method, Green solved the problem of
sheet drawing andbending of notchedbars inthe early 1950s. From
the late 1950s, W. Johnson(1922, working at the University of Cam-
bridge and at UMIST in Manchester) carried out extensive research
work by using the upper bound method for plane-strain forging,
extrusion, rolling and other forming problems. He optimized the
velocity eld by assuming proper variables. In the case of extrusion
K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454 1451
Fig. 38. Velocity eld for upper bound method used by Green (1951) and comparison of obtained pressure with slip-line eld solution.
shown in Fig. 39 (Johnson and Mellor, 1973), the angle is taken as
the parameter for optimization.
Kudo (1960) proposed a method for applying the upper bound
method to axi-symmetric forging and extrusion. He divided the
axi-symmetric billet into several hypothetical units, and derived
mathematical expressions for the velocity of each unit by satisfy-
ing the condition of volume constancy and the requirement that
surface velocities be consistent with those of neighbouring units
and the tool surfaces as demonstrated in Fig. 40.
Fig. 39. Velocity eld used for optimization by Johnson and Mellor (1973).
Fig. 40. Velocity eld for axi-symmetric extrusion by Kudo (1960).
In the 1960s, when cold forging of steel was increasingly used
in the automotive industry, prediction of forming pressure became
a very important subject to avoid fracture of the expensive tools.
The upper bound method for axi-symmetric deformation appeared
just in time and was used extensively in the cold forging industry.
Kudo showed various examples of the use of his method in axi-
symmetric forging and extrusion in the book written with Johnson
(Johnson and Kudo, 1962). In the 1960s and 70s, his method led to
much research activity aimed at nding newtypes of velocity eld,
for instance by Kobayashi (1964), Avitzur (1968) and many others.
In the 1970s, the upper bound method was expanded to three
dimensional problems by Yang and Lee (1978) and others, and then
it was combined with the nite element method, and grew up as
the rigid-plastic nite element method as will be explained in the
later chapter.
7.2. H. Kudo
Hideaki Kudo (19242001) (Fig. 41) graduated from Tokyo
Imperial University in 1945 just after World War II, and started
Fig. 41. H. Kudo.
1452 K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454
his career at the Institute of Science and Technology in Tokyo Uni-
versity under the guidance of S. Fukui. He developedaxi-symmetric
analysis as an approximate energy method without knowing that
his method was related to the upper bound theorem. He was
awarded his doctoral degree for a thesis of the analysis of forging.
From1959 to 1960, Kudo stayed in Hannover THwith O. Kienzle
and in Manchester University with W. Johnson and wrote papers
as well as a book on the upper bound method.
In1960, he joinedthe Government Mechanical Engineering Lab-
oratory in Tokyo as the team leader for cold forging. He worked
hard for the promotion of the industrialization of cold forging and
published many papers and solved various practical problems.
In 1966, Kudo was appointed to a professorship at Yokohama
National University where he held the chair for metal forming until
1989. He carried out fundamental studies of slip-line elds, lubri-
cation, material properties and so on as well as unique forming
methods such as tension-aided can extrusion.
Kudo is one of the originators of the Japanese Society for the
Technology of Plasticity and was President of JSTP for 1985/6. He
alsooriginatedthe International Conference for Technology of Plas-
ticity (ICTP) and served as the chairman of the rst meeting held in
Tokyo in 1984.
8. Finite element method
8.1. Elasticplastic nite element method
The nite element method(FEM) was developedfor elastic anal-
ysis of airplane structures in the 1950s. In this method, a plate was
divided into many hypothetical elements, and equations of equi-
librium at nodal points were developed. Since the equations were
linear in terms of the nodal displacements, they could be solved
with the matrix method using digital computers, which had just
became useful in some limited research facilities in the USA.
O.C. Zienkiwitcz and Y.K. Cheung published a book entitled
The nite element method in structural and continuum mechan-
ics (Zienkeiwicz and Cheung, 1967), in which the method was
explained in detail with the software written in the FORTRAN lan-
guage. Referring to this book, many groups in the world began to
developsoftwareusingdigital computers whichhadbecomeusable
in many countries.
The elasticplastic FEM was developed as an extension of
elastic FEM. Marcal and King (1967) published a paper with an
elasticplastic analysis of a plate specimen with a hole in which
the development of a plastic zone was given as shown in Fig. 42.
They employed a stepwise computation to followthe deformation.
The nodal coordinates and the components of stress and strain in
the element were renewed after each step calculation by adding
the increments to the values before the step. Next year, Yamada
et al. (1968) presented a paper on the stressstrain matrix for
elasticplastic analysis.
This method had a signicant impact on researchers in plastic-
ity and soon, many papers on the elasticplastic analysis of metal
forming problems began to appear. Fig. 43 is the result of analysis
of the initial state of hydrostatic extrusion by Iwata et al. (1972).
In spite of the great potential of the method, it was found that the
calculation error accumulated as plastic deformation proceeds.
In this method, the small deformation formulation, the stress
value is renewed at the end of each step of the calculation as:

A
x
=
B
x
+
x
(18)
where
A
x
and
B
x
are the stresses after and before the step com-
putation, and
x
is the incremental stress during the step. When
an element rotates during the step, the stress state
B
x
xed to the
element also rotates as shown in Fig. 44. This means that Eq. (18)
Fig. 42. Extensionof plastic zone ina plate specimenwitha hole analyzedby Marcal
and King (1967).
Fig. 43. Elasticplastic analysis of hydrostatic extrusion by Iwata et al. (1972).
Fig. 44. Rotation and deformation of an element during a step computation.
K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454 1453
cannot hold without modifying
B
x
, because
A
x
and
B
x
are dened
in different coordinate systems.
To solve this problem, a formulation for large elasticplastic
deformation was put forward by McMeeking and Rice (1975).
It is not known well that Kitagawa and Tomita (1974) ana-
lyzed a large elasticplastic deformation problem in a paper
published earlier in 1974 due to the paper written in the
Japanese language. In the 1980s, commercial software for
elasticplastic deformation appeared, and with the splen-
did increase of computing speed, the method began to
be used in the industry from around 1990. It should be
noted that very small time steps are needed even with
this formulation to avoid error accumulation, and thus
elasticplastic analysis requires very long computing times
even now.
8.2. Rigid-plastic nite element method
Hayes and Marcal (1967) presented a paper on the usage of the
FEMfor optimizing the upper bound method of a plane stress prob-
lem. In the case of the plane stress problem, the stress state could
be calculated from the optimized velocity eld, but this was not
true for other cases.
Anormal stress component canbe decomposedintoa deviatoric
stress component

x
=
x

m
anda hydrostatic stress component

m
=
_

x
+
y
+
z
_
/3. Although the deviatoric stress is related to
the strain rate by Eq. (3), the hydrostatic component is not. For
example,
x
may be written as:

x
=

x
+
m
=
2
3
Y

x
+
m
(19)
where
m
is left indeterminate with the strain rate associated
with the optimized velocity.
If the hydrostatic stress could be determined, this method was
expected to have a great potential. Because the stress could be
calculated at each step without error accumulation, a drastically
shorter computing time than the elasticplastic FEM was possible
although a non-linear problem must be solved by an optimiza-
tion.
Lee and Kobayashi (1973) and Lung and Mahrenholtz (1973)
published papers that enabled stress calculation in the rigid-plastic
analysis. Their theoretical basis was the variational principle with
a Lagrange multiplier, which had been presented in the book
by Washizu (1968). This principle states that when the rigid-
plastic problemis optimizedwiththe Lagrange multiplier tohandle
volume constancy, the multiplier coincides with the hydrostatic
stress.
In FEM with the above principle, one or more multipliers are
needed for each element to obtain the velocity eld which satis-
es the incompressibility condition. Since the Lagrange multipliers
increases the number of variables, the computation time becomes
very long for large scale problems. K. Mori and K. Osakada devel-
oped a nite element method allowing for slight volume change
without increasing the number of variables. In this method the
hydrostatic stress was calculated directly from the slight volume
change which did not give signicant inuence to the deformation.
Fig. 45 shows a result of rolling simulated with this method (Mori
and Osakada, 1982).
In the 1980s, the rigid-plastic nite element method began to
be used by many researchers in metal forming to solve practical
problems (Osakada, 1980; Kobayashi et al., 1989). Similarly to the
elasticplastic nite element method, the rigid-plastic nite ele-
ment method was expanded to three dimensional problems and
was installed in commercial software, especially for forging in the
late 1980s, and used in industry. Recent developments are summa-
rized in Mori (2002).
Fig. 45. Rolling analysis with rigid-plastic FEM by Mori and Osakada (1982).
9. Concluding remarks
With the fast development of information technology in the last
20 years, the nite element method has become the main tool
of metal forming analysis. To realize more accurate simulation,
detailed research is still needed into various areas related to metal
forming such as anisotropy development and changes of metal-
lurgical and mechanical properties during deformation, inelastic
behaviour in unloading and reloading, lubrication, friction, seizure
and fracture.
It seems to be inevitable that the nite element method will
be used more directly in industry. Simulation software may be
integrated into CAD based systems, in which forming simulation
is carried out directly fromCAD of forming tools. In order to enable
small scale metal forming enterprises to use simulation, low cost
software with simplied operations is required.
Although it is impossible for the author to predict far into the
future, on-line control of metal forming processes with simultane-
ous simulation may become possible by the increased computation
speed and new computing algorithms.
Once the simulation method is well advanced, metal forming
engineers will be able to concentrate on more creative and inno-
vative works, e.g., developments of forming processes for products
with very high dimensional accuracy, forming machines for silent
environments, tool coatings for dry metal forming and thermo-
mechanical processes with low tool pressure and high product
strength, etc.
Acknowledgements
The author would like to express sincere thanks to: Dr. R.
Matsumoto (Osaka University), Dr. M. Otsu (Kumamoto Univer-
sity), Prof. K. Mori (Toyohashi University of Technology), Dr. H.
Utsunomiya (Osaka University), Prof. F. Fujita (Tohoku University),
Dr. H. Furumoto (Mitsubishi Heavy Industries) andProf. T. Ishikawa
(Nagoya University), Z.S. Olesiak(University of Warsaw) for provid-
ing the materials, and Dr. J. Allwood and Mark Carruth (Cambridge
University) and Dr. B. Dodd (University of Reading), for providing
the information and for helping to improve the manuscript.
References
Avitzur, B., 1968. Metal forming: Processes and analysis. McGraw-Hill, New York.
Bell, J., 1984. Mechanics of Solids. Vol. 1: The Experimental Foundations of Solid
Mechanics. SpringerVerlag.
1454 K. Osakada / Journal of Materials Processing Technology 210 (2010) 14361454
Bland, D.R., Ford, H., 1948. The calculationof roll force andtorque incoldstriprolling
with tension. Proc. Inst. Mech. Eng. 159, 144153.
Boas, W., Schmid, E., 1930. ber die Temperatuabhngigkeit der Kristallplastizitt.
Z. Physik. A. 61, 767781.
Bridgman, P.W., 1964. Studies in Large Plastic Flowand Fracture. Harvard University
Press.
Dunn, B., 1897. A photographic impact testing machine for measuring the varying
intensity of an impulsive force. J. Franklin Inst. 144 (5), 321348.
Engel, Z., 1994. Hubers Yield Criterion in Plasticity. AGH, Krakow.
Geiringer, H., 1930. Beit zum Vollstandigen ebenen plastiziotats Problem. In: Proc.
3rd Int. Congr. Appl. Mech., Stockholm. 2, pp. 185190.
Gler, V., Sachs, G., 1927. Das VerhaltenvonAluminiumkrystallenbei Zugversuchen,
I. Geometrische Grundlagen. Z. Physik, A. 41, 103115.
Green, A.P., 1951. The compression of a ductile material between smooth dies. Phil.
Mag. 42, 900918.
Guest, J., 1900. On the strength of ductile materials under combined stress. Phil.
Mag., Fifth Ser. 50, 369413.
Hayes, D.J., Marcal, P.V., 1967. Determination of upper bounds for problems in plane
stress using nite element techniques. Int. J. Mech. Sci. 9, 245251.
Hencky, H., 1923. ber einige statisch bestimmte Flle des Gleichgewichts in plas-
tischen Krpern. Z. Angew. Math. Mech. 3, 241251.
Hencky, H., 1924. Zur Theorie plastischer Deformationen und der hierdurch im
Material hervorgerufenen Nachspannngen. Z. Angew. Math. Mech. 4, 323334.
Hill, R., 1948. A theory of the yielding and plastic ow of anisotropic metals. Proc.
Roy. Soc. (London) A 193, 281297.
Hill, R., 1950. The Mathematical Theory of Plasticity. Clarendon Press, Oxford.
Hopkins, H.G., 1980. Obituary professor William Prager. In. J. Mech. Sci. 22, 393
394.
Itihara, M., 1933. Impact torsion test. The technology report of the Tohoku Imperial
University, Sendai, Japan. 11 (1), 1650.
Iwata, K., Osakada, K., Fujino, S., 1972. Analysis of hydrostatic extrusion by the nite
element method. Trans. ASME, Ser. B (J. Eng. Ind.). 94, 697703.
Johnson, W., Mellor, P.B., 1973. Engineering Plasticity. Van Nostrand, London.
Johnson, W., Sowerbby, R., Haddow, J.B., 1970. Plane-Strain Slip-Line Fields: Theory
and Bibliography. Edward Arnold Ltd., London.
Johnson, W., Kudo, H., 1962. The Mechanics of Metal Extrusion. Manchester Univer-
sity Press.
Kitagawa, H., Tomita, Y., 1974. Analysis of plane elastic-plastic large deformation
problems by incremental type nite element method. Trans. JSME 40 (331),
663670 (in Japanese).
Kobayashi, S., 1964. Upper boundsolutionof axisymmetric formingproblems. Trans.
ASME 86, 112126.
Kobayashi, S., Oh, S.-I., Altan, T., 1989. Metal FormingandtheFinite-Element Method.
Oxford University Press.
Kolsky, H., 1949. An investigation of the mechanical properties of materials at very
high rates of loading. Proc. Ro. Soc. (London), B 62, 676700.
Kudo, H., 1960. Some analytical and experimental studies of axi-symmetric cold
forging and extrusion. Int. J. Mech. Sci. 2, 102127.
Lange, K., 1985. Handbook of Metal Forming. McGraw-Hill.
Lee, C.H., Kobayashi, S., 1973. New solutions to rigid-plastic deformation problems
using a matrix method. Trans ASME, Ser. B. 95, 865873.
Lippmann, H., Mahrenholts, O., 1967. Plastomechanik der Umformung metalllisher
Werstoff. SpringerVerlag.
Lode, W., 1926. Versuche ber den einu mitteren Hauptspannung auf das Flie
der Metalle Eisen. Kupfer und Nickel. Z. Physik. 36, 913939.
Lung, M., Mahrenholtz, O., 1973. A nite element procedure for analysis of metal
forming processes. Trans. CSME 2, 3136.
Manjoine, M.J., Ndai, A.L., 1940. High-speed tension tests at elevated temperatures.
Proc. Am. Soc. Test. Mater. 40, 822839.
Marcal, P.V., King, I.P., 1967. Elastic-plastic analysis of two-dimensional stress sys-
tems by the nite element method. Int. J. Mech. Sci. 9, 143145.
McMeeking, R.M., Rice, J.R., 1975. Finite-element formulations for problems of large
elastic-plastic deformation. Int. J. Solid Structures 11, 601616.
Mori, K., Osakada, K., 1982. Simulation of plane strain rolling by the rigid-plastic
nite element method. In. J. Mech. Sci. 24, 515527.
Mori, K., 2002. Rigid-plastic niteelement simulationof metal formingprocesses. In:
Kiuchi, M., et al. (Eds.), Advanced Technology of Plasticity 2002, 1, pp. 261270.
Ndai, A., 1931. Plasticity A Mechanics of the Plastic State of Matter.
McGrawHill.
Ndai, A.L., 1937. Plastic behavior of metals in the strain-hardening range. Part I. J.
Appl. Phys. 8, 205213.
Ndai, A.L., 1939. The force required for rolling steel strip under tension. J. Appl.
Mech. 6, A5462.
Olesiak, Z.S., 2000. Professor Maksymilian Tytus Huber his Life and activity. Eur. J.
Mech. A/Solids 19, 101119.
Orowan, E., Pascoe, K.P., 1946. A simple method of calculating roll pressure and
power consumption in hot at rolling. Iron and Steel Inst. Special Report No. 34,
124146.
Orowan, E., 1943. The calculation of roll pressure in hot and cold at rolling. Proc.
Inst. Mech. Eng. 150, 140167.
Orowan, E., 1934. Zur Kristallplastizitt I: Tieftemperatur-plastizitt undBeckersche
Formel. Z. Physik. A. 89, 605613.
Osakada, K., 1980. A reviewof nite element analysis of metal forming. In: Proc. 4th
Int. Conf. on Production Engineering, Tokyo, pp. 4449.
Polanyi, M., 1934. ber eine Art Gitterstrung, die einen Kristall plastisch machen
knnte. Z. Physik, A. 89, 660664.
Prager, W., Hodge, P.G., 1951. Theory of Perfectly Plastic Solids. Wiley, New York.
Prandtl, L., 1924. Spannungsverteilung in plastischen Krpern. Proc. 1st Int. Congr.
Appl. Mech. Delft., 4354.
Prandtl, L., 1920. ber die Hrte plastischer Krper. Nachrichtenvonder Kniglichen
Gesellschaft der Wissenschaften zu Gttingen. Math. -physik. Kl., 7485.
Reuss, A., 1930. Bercksichtigung de elastishcen Formnderng in der Plastizitt-
slehle. Z. Angew. Math. Mech. 10, 266274.
Sachs, G., 1928. Zur Ableitung einer Fliebedingung. Z. Vereins Deutscher Inge-
nieure. 72, 734736.
Sachs, G., 1927. Zur Theorie des Ziehvorganges. Z. Angew. Math. Mech. 7, 235236.
Schmid, E., 1926. ber die SchubverfestigungvonEinkristallenbei plastischer Defor-
mation. Z. Physik. A 40, 5474.
Hopkins, H.G., Sewell, M., 1982. Rodney Hill, Biographical Note. In: Hopkins, H.G.,
Sewell, M. (Eds.), Mechanics of Solids The Rodney Hill 60th Anniversary Volume.
Pergamon Press Ltd., pp. 921.
Shima, S., Oyane, M., 1976. Plasticity theory for porous metals. Int. J. Mech. Sci. 88
(6), 285291.
Siebel, E., 1923. Untersuchungen ber bildsame Formnderung unter besonderer
Bercksichtigung des Schmiedens. Maschinenbau/Betrieb. 9, 307312.
Sokolovskii, V.V., 1948. The theory of plasticity outline of work done in Russia.
Trans. American Soc. Mech. Eng. (Trans ASME). J. Appl. Mech. 13, A1A9.
Taylor, G.I., Elam, C.F., 1923. The distortion of an aluminium crystal during a tensile
test. Proc. Roy. Soc. (London) A 102, 643647.
Taylor, G.I., 1938. Plastic strain in metals. J. Inst. Metals 62, 307324.
Taylor, G.I., 1934. The mechanism of plastic deformation of crystals, Part I Theo-
retical. Proc. Roy. Soc. (London) A 145, 362404.
Taylor, G.I., Quinney, H., 1932. The plastic distortion of metals. Phil. Trans. Roy. Soc.
(London) 230, 323362.
Timoshenko, S.P., 1953. History of Strength of Materials. Dover Publication, New
York.
Trinks, W., Hitchcock, J.H., 1935. Deformation of Rolls during Cold Rolling. Roll Neck
Bearings. ASME Research Publications, New York, pp. 3341.
Trinks, W., 1937. Pressure and roll attening in cold rolling. Blast Furnace Steel Plant
25, 617623.
von Krmn, Th., 1925. Beitrag zur Theorie des Walzvorges. Z. Angew. Math. Mech.
5, 139141.
von Krmn, Th., 1967. Wind and Beyond. Little Brown & Inc.
von Mises, R., 1913. Mechanik der Krper im plastische-deformablen Zustand
Nachrichten von der Kniglichen Gesellschaft der Wissenschaften zu Gttingen.
Math. -physik. Kl., 582592.
Washizu, K., 1968. Variational Methods in Elasticity and Plasticity. Pergamon Press.
Yamada, Y., Yoshimura, N., Sakurai, T., 1968. Plastic stressstrainmatrixandits appli-
cation for the solution of elastic-plastic problems by nite element method. Int.
J. Mech. Sci. 10, 343354.
Yamaguchi, K., 1929. Slip-bands of compressed aluminium crystals. Part I Dis-
tortion by single slipping and a tentative theory of work-hardening of metals.
Scientic Papers of the Institute of Physical and Chemical Research, 11 (205),
223241.
Yang, D.Y., Lee, C.H., 1978. Analysis of three-dimensional extrusion of sections
through curved dies by conformal transformation. Int. J. Mech. Sci. 26, 541552.
Zienkeiwicz, O.C., Cheung, Y.C., 1967. The Finite Element Method in Structural and
Continuum Mechanics. Mc-Graw-Hill, New York.